Commodity Market Forecasting What know and what we don t know about the future Global Insights 40 th Anniversary 22 October 2003 3001 Becker Road, P.O. Box 1269 Langley, WA 98260, USA Tel: +1+360 321 4707 Fax: +1+360 321 4709 31 Mt Pleasant, London WC1X OAD, UK Tel: +44+20 7278 7788 Fax: +44+20+7278 0003 Additional offices and representatives: Philadephia (Exton), Rio de Janiero and Sydney
Contents 1) The Lessons of History 2) Commodity Markets Today 3) Key Issues for the Future Slide 2
There are two incompatible approaches to commodity price forecasting Fundamental Demand is derived from economic activity Supply takes place in response to resource availability Imbalances are resolved by inventory changes which drive prices Used for both long-term and short-term forecasts Technical All relevant market information is contained in the price Price trends are the key to price forecasting Prices forecast by purely statistical techniques and chart analysis Used mostly for very short-term forecasts Slide 3
Commodity futures prices are no substitute for a decent forecast Example: average absolute error using LME 3 month aluminum price as a forecast is $100/t this is 3-4 times the average absolute error of forecasts by Chase Econometrics/ /CRU Future price is actually today s price for a physical delivery in the future This incorporates today s market expectations new information obviously affects expectations in the mean time; so outcomes are not as predicted But: futures prices happen to be a statistically unbiased predictor of actual prices Slide 4
Three equally unbiased forecasts but one is clearly better! Price 120 110 Right direction 75% of the time 100 90 80 1 2 3 4 5 Forecast A Forecast B Forecast C Actual Slide 5
Statistical trend analysis and fundamental research yield similar results for copper Real Copper Price 7000 6000 Statistical trend Median - LRMC 5000 4000 3000 2000 SRMC 1000 Jan-80 Jan-81 Jan-82 Jan-83 Jan-84 Jan-85 Jan-86 Jan-87 Jan-88 Jan-89 Jan-90 Jan-91 Jan-92 Jan-93 Jan-94 Jan-95 Jan-96 Jan-97 Jan-98 Jan-99 Jan-00 Jan-01 Jan-02 Jan-03 Slide 6
30000 25000 20000 15000 10000 5000 0 The same is true for nickel Real Nickel Price Median - LRMC Statistical trend SRMC Jan-83 Jan-84 Jan-85 Jan-86 Jan-87 Jan-88 Jan-89 Jan-90 Jan-91 Jan-92 Jan-93 Jan-94 Jan-95 Jan-96 Jan-97 Jan-98 Jan-99 Jan-00 Jan-01 Jan-02 Jan-03 Jan-82 Slide 7 Jan-80 Jan-81
But the aluminum trend is improbably low due to structural shift in power markets Real Aluminum Price 4500 4000 Median - LRMC 3500 Statistical trend 3000 2500 2000 1500 1000 SRMC Jan-80 Jan-81 Jan-82 Jan-83 Jan-84 Jan-85 Jan-86 Jan-87 Jan-88 Jan-89 Jan-90 Jan-91 Jan-92 Jan-93 Jan-94 Jan-95 Jan-96 Jan-97 Jan-98 Jan-99 Jan-00 Jan-01 Jan-02 Jan-03 Slide 8
Similar result in zinc, probably due to Europe exit barriers & Chinese competition Real Zinc Price 3000 Statistical trend 2500 Median - LRMC 2000 1500 1000 SRMC 500 Jan-80 Jan-81 Jan-82 Jan-83 Jan-84 Jan-85 Jan-86 Jan-87 Jan-88 Jan-89 Jan-90 Jan-91 Jan-92 Jan-93 Jan-94 Jan-95 Jan-96 Jan-97 Jan-98 Jan-99 Jan-00 Jan-01 Jan-02 Jan-03 Slide 9
The really important decisions tend to require fundamental forecasts New project feasibility studies are not bankable without a fundamental forecast Long-term fundamental forecasts are required for asset valuations and for effective implementation of value based management Critical technical issues (definition of reserves, life of mine planning) require fundamental forecasts Industry stakeholders continue to make or buy fundamental forecasts despite their mediocre accuracy record Slide 10
Forecasting methods have adapted to the market environment Period Market Environment Forecasting Trends 1960-1973 1973-1982 1982-1989 1989-2001 Strong demand growth, nationalization to secure economic rents, stable or rising prices Oil crisis, severe cycles in demand and price, threat of hyper-inflation Reduced inflation, high interest rates, weak demand, sharp fall in prices, privatization begins Collapse of communism, sustained growth based on globalization and low inflation, prices still fall First formal econometric forecasting models; classic market clearing; aggregated data Highly disaggregated forecasting linked to detailed macroeconomic models Focus on supply side forecasting and understanding production costs Hybrid forecasting techniques geared to understanding structural issues Slide 11
The market clearing model a simplified version World Economy Production Costs Demand Supply Market Balance Inventories Operating Rate Prices Capacity Slide 12
The demand side lessons World Economy Demand Far richer macro forecasts but more detailed is not necessarily more accurate Market Balance Minimal progress on understanding price elasticities Cost culture has weakened link between demand and investment Prices Capacity Slide 13
The supply side lessons Comprehensive mine by mine, plant by plant databases Sophisticated understanding of cost structures World Economy Supply Production Costs Market Balance Investment dynamics have experienced structural change Operating Rate Prices Capacity Slide 14
Commodity markets have asymmetrical pricing dynamics Condition Required Signals Price Implication Surplus markets Deficit markets In between markets Demand is fully satisfied and stocks are high and rising; market signals the need to cut production All available capacity is operating but stocks are low and falling; auction market until marginal consumer defers demand Demand is satisfied and producers who wish to operate are doing so; inventories are reasonable and/or trends are unclear Prices fall to the industry s short-run marginal costs (SRMC) Prices rise to large premium over costs of even the highest cost producer; reflect shortrun opportunity cost of the marginal consumer Prices reflect SRMC plus a premium based on future expectations; in equilibrium this equals industry long-run marginal costs (LRMC) Slide 15
Prices are more often predictably below trend than unpredictably above Metal Price Deviations from Trend January 1980 to July 2003 Average of 5 Metals Al Cu Pb Zn Ni Monthly Data Below Average # 142 193 191 147 140 163 Average Deviation -12.2% -19.9% -19.4% -11.8% -18.5% -16.4% Max Deviation -28.7% -40.1% -40.0% -33.1% -43.4% -37.1% Above Average # 141 90 92 136 143 120 Average Deviation 23.3% 16.6% 22.9% 22.1% 34.2% 23.8% Max Deviation 106.7% 55.5% 122.8% 98.6% 191.7% 115.1% Annual Data Below Average # 11 13 15 13 12 13 Average Deviation -10.9% -22.8% -19.2% -8.7% -16.6% -15.6% Max Deviation -21.9% -35.5% -35.2% -29.8% -32.8% -31.0% Above Average # 12 10 8 10 11 10 Average Deviation 21.0% 10.9% 20.7% 23.5% 34.5% 22.1% Max Deviation 70.4% 29.6% 76.2% 68.0% 104.3% 69.7% Slide 16
2600 2400 Fan Charts illustrate the role of expectations in determining prices Structure of Forward Aluminum Prices Since 1989 2200 2000 1800 1600 1400 1200 1000-1099 1100-1199 1200-1299 1300-1399 1400-1499 1500-1599 1600-1699 1700-1799 1800-1899 1900-1999 2000-2099 2100-2199 2200-2299 2300-1000 Cash 3- month 15- month 27- month 39 month 54- Month Slide 17
Contents 1) The Lessons of History 2) Commodity Markets Today 3) Key Issues for the Future Slide 18
The big picture across commodity metals reflects the global economy Commodity metal prices peaked between Q3 2000 and Q1 2001 reflecting a global slowdown that predated 9/11 The period of decline lasted 18-24 months basically flat to recovering since mid 2002 The recovery has been very uneven to date well over 50% of volume growth in most metals is coming from China most of the rest from other Asia and North America CIS has turned round, but absolute volumes are low Slide 19
Prices, however, are still well below pre- recession averages except nickel 40% Average 3-month price: June 2003 vs 1993-2002 30% 20% 10% Special case reflecting the failure of new technology which deterred investment in late 1990s 0% -10% -20% -30% Copper Lead Zinc Aluminium Nickel Tin Slide 20
Normal (inverse) relationships between inventories and prices except for nickel LME stock and price changes, Dec 02 June 03 20% 10% 0% -10% -20% -30% LME stocks LME prices Zinc Aluminium Lead Copper Tin Nickel Slide 21
The heatchart heatchart shows considerable divergence among metals CRU Heatchart TM 2003: Nickel hottest Surplus/deficit to stocks ratio 40% 30% 20% 10% 0% -10% -20% Aluminium Lead Nickel Zinc Tin Copper -30% 50% 75% 100% 125% 150% 175% Stock:consumption ratio versus 15 year average Slide 22
We forecast that there will be relatively few changes in 2004 Surplus/deficit to stocks ratio 20% 10% 0% -10% -20% -30% -40% CRU Heatchart TM 2004 Forecast Aluminium Zinc Lead Nickel Tin Copper -50% 50% 75% 100% 125% 150% Stock:consumption ratio versus 15 year average Slide 23
Contents 1) The Lessons of History 2) Commodity Markets Today 3) Key Issues for the Future Slide 24
There are four major issues that we can now identify The impact of China on demand and supply Globalization versus protectionism The ability to manage the cycle and deliver shareholder value The impact of new technologies Slide 25
Chinese risk is now a critical factor in virtually every commodity market China accounts for 50%-70% of the annual growth in global demand (greater impact than Japan inn the 1960s) will coastal prosperity spread to interior China? when will the growth rate mature? China is a major producer and exporter of key commodities most ferroalloys, zinc, magnesium etc unsustainable development of power-intensive sector Potential exists for unexpected and severe swings in net trade balances Slide 26
The triumph of globalization over protectionism is not a done deal Steel sectors continue to be heavily protected will be increasingly challenged by technology change in upstream iron and steel sectors Non-ferrous smelting and refining is being subsidized via tariffs on imports, especially in Asia copying traditional Japanese practices which contribute to excess capacity bias Emerging problem of deindustrialization starting to affect highly efficient, competitive downstream sectors in North America will be a problem in Europe in next 5-10 years Slide 27
Good news and bad news on cyclical management potential Industry consolidation is producing stronger and more disciplined corporate entities Alcoa, Alcan, Rusal etc in aluminum Diversified companies like Rio Tinto, BHP-Billiton, Anglo American and Xstrata in other markets Governments have basically exited these sectors Potential loss of North American swing capacity reduces overall industry supply flexibility devastating experience of zinc (no US swing capacity) in the current cycle virtually no progress in Europe towards flexible production concepts Slide 28
New technologies potentially trigger structural shifts that disrupt markets Inert anode technology in aluminum potential drop of 25% in real cost of producing aluminum Leaching of primary copper sulphide deposits fundamental reduction in costs, scale and entry barriers Similar developments in other markets Mt Isa s zinc technology, APL in nickel etc New technologies for primary iron production Rio s Hismelt project and its rivals threaten integration of iron and steelmaking Slide 29
The challenges have escalated at least as fast as the lessons have been learned! Global and national economic risk Political, social and environmental constraints The Renaissance Man challenge Technology change Industry management issues and challenges Slide 30